Re: [PATCH] iommu/mtk_iommu: Use devm_kcalloc() instead of devm_kzalloc()

From: AngeloGioacchino Del Regno
Date: Mon Feb 12 2024 - 03:49:37 EST


Il 11/02/24 19:22, Erick Archer ha scritto:
This is an effort to get rid of all multiplications from allocation
functions in order to prevent integer overflows [1].

Here the multiplication is obviously safe because MTK_PROTECT_PA_ALIGN
is defined as a literal value of 256 or 128.

For the "mtk_iommu.c" file: 256
For the "mtk_iommu_v1.c" file: 128

However, using devm_kcalloc() is more appropriate [2] and improves
readability. This patch has no effect on runtime behavior.

Link: https://github.com/KSPP/linux/issues/162 [1]
Link: https://www.kernel.org/doc/html/next/process/deprecated.html#open-coded-arithmetic-in-allocator-arguments [2]
Signed-off-by: Erick Archer <erick.archer@xxxxxxx>

Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@xxxxxxxxxxxxx>

---
drivers/iommu/mtk_iommu.c | 2 +-
drivers/iommu/mtk_iommu_v1.c | 4 ++--
2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/iommu/mtk_iommu.c b/drivers/iommu/mtk_iommu.c
index 7abe9e85a570..9aae6eb604b1 100644
--- a/drivers/iommu/mtk_iommu.c
+++ b/drivers/iommu/mtk_iommu.c
@@ -1264,7 +1264,7 @@ static int mtk_iommu_probe(struct platform_device *pdev)
data->plat_data = of_device_get_match_data(dev);

/* Protect memory. HW will access here while translation fault.*/
- protect = devm_kzalloc(dev, MTK_PROTECT_PA_ALIGN * 2, GFP_KERNEL);
+ protect = devm_kcalloc(dev, 2, MTK_PROTECT_PA_ALIGN, GFP_KERNEL);
if (!protect)
return -ENOMEM;
data->protect_base = ALIGN(virt_to_phys(protect), MTK_PROTECT_PA_ALIGN);
diff --git a/drivers/iommu/mtk_iommu_v1.c b/drivers/iommu/mtk_iommu_v1.c
index 25b41222abae..45cd845d153f 100644
--- a/drivers/iommu/mtk_iommu_v1.c
+++ b/drivers/iommu/mtk_iommu_v1.c
@@ -621,8 +621,8 @@ static int mtk_iommu_v1_probe(struct platform_device *pdev)
data->dev = dev;

/* Protect memory. HW will access here while translation fault.*/
- protect = devm_kzalloc(dev, MTK_PROTECT_PA_ALIGN * 2,
- GFP_KERNEL | GFP_DMA);
+ protect = devm_kcalloc(dev, 2, MTK_PROTECT_PA_ALIGN,
+ GFP_KERNEL | GFP_DMA);
if (!protect)
return -ENOMEM;
data->protect_base = ALIGN(virt_to_phys(protect), MTK_PROTECT_PA_ALIGN);
--
2.25.1